中文版成瘾严重性指数量表在参加美沙酮维持治疗的海洛因成瘾者中的信度和效度评价

RELIABILITY AND VALIDITY OF ADDICTION SEVERITY INDEX (ASI) CHINESE VERSION IN EVALUATION OF HEROIN ADDICTS ON METHADONE MAINTENANCE TREATMENT

背景:为遏制艾滋病病毒在海洛因成瘾者中的传播,我国于2004年开始选用美沙酮口服液作为海洛因的替代药物,对海洛因成瘾者进行长期维持治疗。为更好地评价美沙酮维持治疗(methadone maintenance treatment,MMT)的效果,研究者引进和翻译了在国外广泛用于评价药物成瘾的一种工具《成瘾严重性指数量表》(Addiction Severity Index,ASI)第5版。目的:评价中文版ASI(ASI-C)在接受MMT的海洛因成瘾者中应用的信度和效度。方法:对1005名(男744名,女261名)MMT受治者用ASI-C进行面对面访谈,其中35名受治者接受了间隔7d的重复测量,分析ASI的信度和效度。结果:ASI-C与原量表具有相近的结构,通过因子分析,有7个因子的特征值大于1,ASI-C萃取出7个维度;通过与相关效度条目的检验,ASI-C具有较好的区分效度,各效度检验条目与维度评分间相关系数在0.09-0.80。内部一致性信度测量显示平均Cronbach’α为0.68(0.52-0.88),有5个维度Cronbach’α大于0.60。相隔7d的重测信度达到可接受的水平。结论:ASI-C在参加美沙酮维持治疗的海洛因成瘾者中应用,具有较好的结构效度和区分效度,5个维度的内部一致性信度Cronbach’α达到可接受的水平,2维度上还有待提高。初步测量的重测信度达到可接受水平。

Background：In order to improve the management of opiate abuse problem and contain the spread of HIV, China established methadone maintenance treatment （MMT） programs in early 2004. In order to evaluate the effectiveness of MMT, we translated the 5th edition of Addiction Severity Index （ASI） into Chinese as a measurement instrument. Objective：To evaluate the reliability and validity of the 5th edition ASI, Chinese version （ASI-C） in heroin addicts receiving MMT in China. Methods：One thousand and five heroin addicts receiving MMT were recruited by convenient sampling, 744 men and 261 women. They were all interviewed with the ASI-C questionnaire, and 35 were re-interviewed at an interval of seven days to assess test-retest reliability. Results：Factor analysis for construct validity, with eigenvalue 〉1, seven factors were extracted. Comparing the 7 domains with the criterion indictors in ASI-C, it showed good discriminant validity （r=0.09-0.80）. The average of the Cronbach’s alpha for internal consistency of CSs was 0.68, from 0.52 to 0.88, five domains’ Cronbach’s alpha were above 0.60. Test-retest reliability of composite scores （CSs） of ASI-C were good enough with interval 7days overall. Conclusion：ASI-C in drug users had good enough construct and discriminant validity. For interval consistency of Cronbach’s alpha, five domains in ASI-C were acceptable, but two domains’ should be improved. Test-retest reliability was acceptable overall.